Streaming Media

 
Streaming Media on Facebook Streaming Media on Twitter Streaming Media on LinkedIn Streaming Media on YouTube
Sponsors

IBM Introduces AI-Driven Captioning for Live and On-Demand Video
Watson Captioning will learn over time, becoming more accurate as users make edits and add their own custom glossaries.
Learn more about the companies mentioned in this article in the Sourcebook:
{0}

In an attempt to make adding closed captions to live and recorded video less of a chore, IBM introduced Watson Captioning today. A standalone service, it uses artificial intelligences (AI) to learn over time and allows companies to add their own glossaries of words and phrases.

Watson Captioning was developed for media companies, explains David Kulczar, senior product manager of Watson Video Analytics at IBM Watson Media. It lets them create and edit captions in streamlined workflows, reducing the costs of doing the work by hand, while meeting regulatory requirements. However, IBM also sees uses for it in the education and enterprise IT markets.

While IBM isn't guaranteeing customers a set level of accuracy with Watson Captioning, the service learns new information over time, reducing errors. Companies can help the process by creating their own custom lists of relevant words and phrases. Using AI, the service looks at video context and the edits made to correct mistakes.

"As we’ve seen a major shift across industries towards video content, companies are faced with new challenges towards meeting costly compliance and production needs," Kulczar says. "Watson Captioning taps IBM’s machine learning technology to meet these needs. It enables media and entertainment companies to produce content that is accessible to all audiences, while empowering production teams and streamlining workflows."

Watson Captioning is a software-as-a-service (SaaS) offering that's part of the IBM Cloud platform, and it's commercially available starting today. It can input any type of video, but will only output in MP4. Pricing is on a tiered system based on how much video is processed, with different prices for live and on-demand content.

Related Articles
Ustream will join other IBM acquisition targets—Clearleap, Cleversafe, and Aspera—in comprising an end-to-end cloud video solution.
The tennis tournament is using an AI-based solution that IBM calls "cognitive highlights" to deliver the most important clips to fans
Streaming Media's Tim Siglin interviews IBM Cloud Video Senior Solutions Engineer Scott Grizzle at Streaming Media West 2017.
Artificial intelligence and machine learning, along with deep learning and neural networks, are solving OTT challenges from encoding quality to closed captioning.